vllm vs vllm-ascend
A neutral, constraint-first comparison - live GitHub stats and typed relationships, not marketing.
| vllm | vllm-ascend | |
|---|---|---|
| Tagline | A high-throughput and memory-efficient inference and serving engine for LLMs | Community maintained hardware plugin for vLLM on Ascend |
| Stars | 86k | 2.4k |
| Forks | 19k | 1.6k |
| Open issues | 5.6k | 2.3k |
| Language | Python | C++ |
| License | Apache-2.0 | Apache-2.0 |
| Last pushed | Jul 7, 2026 | Jul 7, 2026 |
| Categories | Inference & Serving, Model Training | Inference & Serving |
vllm
vLLM is a fast and efficient library designed to serve large language models (LLMs) with high throughput while being mindful of computational resources. It supports various model optimizations, quantization techniques, and offers seamless integration with popular Hugging Face models.
Python
vllm-ascend
A repository providing a community-maintained hardware plugin for the vLLM (a large language model inference framework) specifically tailored for Ascend hardware, facilitating efficient LLM serving and deployment.
C++